Horse Racing Strategies Megathread
-
@james-menhenitt said in ***Horse Racing Shared Strategies Megathread + Downloads***:
@martin Ummm.... I think so! I just exported this:
0_1657111861842_rules_export (1).json
I thought/think it is simply your BTF?
OK so this imported version shows you have entry time set at market open all the way up until 10 minutes before the start of the race. That's why you are getting higher odds as BSP will most likely be lower on most favourites than earlier in the day. You need to set market entry time to about ten seconds before the race not at market open.
This is the result from that race if you do so:
2022-07-04 20:45 Ripon - Matty Too - BSP 3.47 / Entry Price - 3.75 -
@john-folan Which is why I am confused with the difference between the results price (6.4) and the price I got backed at 30seconds before start (3.65)... I'm sure it'll all even out though
-
@ian-parry it is. But. The closer to the off you are the more accurate. Especially with the real low odds stuff. They are normally well and truly settled 1 minute before the official start time. You may get the odd one that changes but not enough to make a difference to the strategy in real terms.
-
@david-walsh said in ***Horse Racing Shared Strategies Megathread + Downloads***:
@james-menhenitt Currently any strategy that involves a favourite must be built using take BSP. It doesn't work if you select a time before the race to place your bet.
If you were to select to place your bet 10 mins out.......
The software looks backwards through history, finds the horse that was the BSP favourite and then records the price of that horse 10 mins out. Then returns that result.
The problem being that 10 minutes out the favourite could have been a different runner. And going forwards it's impossible to know 10 minutes out which horse is going to end up the BSP favourite. So to test any strategy with a favourite and return the same results you need to select BSP, both on your strategy builder and your automation. Hope I've explained this ok to understand.So from this it sounds like the results for some of the strategies in this post (eg. Martins back faves, Keith's back fave) would have incorrect results since they use the favourite rule and specify a time before the race. Is this correct?
-
@david-walsh That's really REALLY helpful, thanks David!!!
-
@david-walsh said in ***Horse Racing Shared Strategies Megathread + Downloads***:
@james-menhenitt Currently any strategy that involves a favourite must be built using take BSP. It doesn't work if you select a time before the race to place your bet.
If you were to select to place your bet 10 mins out.......
The software looks backwards through history, finds the horse that was the BSP favourite and then records the price of that horse 10 mins out. Then returns that result.
The problem being that 10 minutes out the favourite could have been a different runner. And going forwards it's impossible to know 10 minutes out which horse is going to end up the BSP favourite. So to test any strategy with a favourite and return the same results you need to select BSP, both on your strategy builder and your automation. Hope I've explained this ok to understand.@Adam is adding this to a future update where you get the fav 10 mins out, at the moment I do the filter as close to the off as possible for accurate results
-
@david-walsh The other rules you see on there are just my own settings for other things. Ignore those
-
-
@james-menhenitt
For your rule to back the favourite at BSP it should look like the screenshot. I have the timing to run 30 secs out to 15 mins after the start time. Then in your conditions apply a "market not in play" rule. ! Hope this helps 0_1657113855706_Screenshot betangel.png -
@david-walsh Thanks, David - I do get the principle completely and yes, that would be the difference between BSP and mine if I set it x seconds or x minutes before. So I appreciate I might need to tweak the BetAngel automation.
The reason why I posted last night was while I could understand a gap between 3.47 and 3.65 in the 30s before the start time and BSP being down to market movements in that time, the strategy recorded 6.4 in the results which is a very large spread that I'm not convinced could be down to the timing difference. Obviously I'm messing something up somewhere. I'm going to change the automation tonight to BSP if I can figure that out in Guardian - see what happens!
Thanks again
-
@martin Ummm.... I think so! I just exported this:
0_1657111861842_rules_export (1).json
I thought/think it is simply your BTF?
-
@james-menhenitt Currently any strategy that involves a favourite must be built using take BSP. It doesn't work if you select a time before the race to place your bet.
If you were to select to place your bet 10 mins out.......
The software looks backwards through history, finds the horse that was the BSP favourite and then records the price of that horse 10 mins out. Then returns that result.
The problem being that 10 minutes out the favourite could have been a different runner. And going forwards it's impossible to know 10 minutes out which horse is going to end up the BSP favourite. So to test any strategy with a favourite and return the same results you need to select BSP, both on your strategy builder and your automation. Hope I've explained this ok to understand. -
@james-menhenitt can you share the strategy so I can take a look?
-
@david-walsh So the latest (having not yet made David's suggested change) is the setting is currently 30s before the expected start time.
This has meant that I'm matching far more with the results that the software is recording.
However, for example the 20.45 at Ripon, my automation placed a back 30s out from the start and it was Matty Too at 3.65.
Looking at the software results, the BSP was 3.47. Fine. But the Entry Price is showing as a whopping 6.4! Resulting obviously in a significant difference in the profit I've recorded and the profit the strategy is recording.
Is there a way of knowing what time/how long before start the software is scraping the entry price? There's a big difference in price which suggests the times must be reasonably far apart....
Any suggestions welcome!!
-
@david-walsh Thanks David, I'll give that a try tonight!
-
@james-menhenitt You need to change your automation to "Place Back SP bet"
in Position 1 (sorted in order of favouritism) -
@paul-foster couple of things I really like here.
- It's essentially a base filter that you haven't tried to improve yet.
- Not huge roi but a steady profit month to month, very low variance!